タグ：proxy

Local OpenAI/Anthropic/Responses-compatible proxy that uses TypeSafe Jev to route model and thinking level after protocol, context, and quota filtering; records each turn.

Shift every LLM call to the cheapest model that can handle it. Routing decided by TypeSafe Jev in ~180 ms. No training data. Policy in plain YAML. TypeScript and Python.

Local proxy for the TypeSafe Jev API: records, caches and replays calls, with a web UI to inspect them. Unofficial.